Prediction of Introduced Distortions Parameters in Lossy Image Compression

S. Krivenko (1) , V. Lukin (1) , B. Vozel (2)

1 Kharkov National University
2 IETR - Institut d'Électronique et des Technologies du numéRique

The paper addresses a problem of predicting parameters of distortion introduced by lossy image compression. Such a prediction is needed to produce a desired quality of compressed images quickly. It is shown that mean square error and peak signal-to-noise ratio can be predicted with high accuracy if images subject to compression are practically noise-free an with appropriate accuracy if images subject to compression are noisy. Approaches to improve accuracy of parameter prediction in compressing of noisy images are proposed and compared. Practical recommendations are given. © 2018 IEEE.

Keywords: Image compression; noise; prediction; Image quality

Domains: Engineering Sciences [physics]

S. Krivenko, V. Lukin, B. Vozel. Prediction of Introduced Distortions Parameters in Lossy Image Compression. 2018 International Scientific-Practical Conference on Problems of Infocommunications Science and Technology, PIC S and T 2018, Oct 2018, Kharkov, Ukraine. pp.447-451, ⟨10.1109/INFOCOMMST.2018.8632130⟩. ⟨hal-02089123⟩